by John G. Dove (Consultant, Paloma & Associates)
There’s a Problem of Inertia in Open Access
Great strides have been made to pursue the objective of having scholars worldwide have unfettered access to the body of work that represents the state of knowledge in each field of inquiry. There are now a good number of peer-reviewed journals which are open access from day one. Almost all of the top scholarly journal publishers worldwide have open access journals of their own. Just about every academic discipline has at least one good open access journal published in that field. And almost all publishers acknowledge that the lead authors of an article can, by right, (sometimes after an embargo period) share a version of their article in an archive or website that provides open access to that version.
Some of the most important funding organizations in the world have mandates that research which they fund must be reported on in open access journals (again, sometimes after an embargo period). Now more than 100 universities worldwide have adopted open access policies strongly encouraging scholars at their universities to post the results of their scholarly work in an institutional repository.
The success of the Open Access efforts over the past 15 years can be seen by looking at the list of members of the Open Access Scholarly Publishers Association. http://oaspa.org/membership/members/ It includes almost every one of the scholarly publishers in the world as well as the leading copyright rights management organization (CCC).
However, even with mandates, policies, and endorsement from publishers, only a small percentage of scholarly journal articles worldwide are ending up open to all scholars to read. It’s apparently the case that wherever there are mandates or policies in place a sizeable adoption of open access is achieved, but this represents only a small portion of all scholarly work and does not cover research that was done before the mandates or policies were in place.
In places where there are no mandates or policies (in other words the vast majority of journal articles) the percentages of articles available in an open source is very low. Some put it as low as 5% for some disciplines.
Here’s a little exercise which I’ve now done looking at research papers in a wide variety of disciplines. Look at the referenced sources in a recently published paper. Unless you are reading this paper at one of the few fully-funded research libraries, you will find that a significant number of the referenced sources are unavailable to you. Open access is simply not there. If you talk to any scholar or would-be scholar in any discipline about how they go about a careful reading of an important paper in their field you will find that a central part of the reader’s experience is to browse through the full-text of several of the referenced sources. Very often these are papers written well before a mandate or policy on Open Access was in place in their discipline. If the reader is at a less than fully funded college or university, or is reading the article from a less developed country, an attempt to do a detailed study of an article is a throw-back to days before computers. Lots of the referenced sources will have to be obtained by inter-library loan or not at all. Your ability to participate in the scholarly inquiries of your field are highly constrained.
How can the vast majority of scholarly articles that represent the basis for scholarly work make their way into an accessible place so that the mission of Open Access can be accomplished? It should not surprise anyone that there is inertia among scholars to exercise their right to share. Scholars are busy. It generally takes an active step by authors to share. Of course, publishers could open up access after some period of time and some may do so, but there’s no obvious self-interest among commercial publishers that would drive them to do this.
Publishers are Missing a Trick
I think there’s a case to be made that journal publishers may be missing a trick. There is a point in time when a publisher’s self-interest in the quality of their about-to-be-published work would be well-served by encouraging authors of referenced sources to share their past articles. This is also a moment in time at which the authors of referenced sources are also missing a trick but are unaware of it.
Imagine the publisher of an article about to be published. They could examine the soon-to-be-published article and take note of the cited sources in that article. Which of them are originally published in an open access journal or have had a version of the article archived in an institutional repository? The utility of this about-to-be-published article is clearly enhanced by having as many referenced sources as possible be eventually made open. That way readers can fully absorb the import of this article as they can browse the underlying research on which the new results are based. [Nothing in what I propose should ever imply that selection of which articles to cite should be based on whether they are open or not. Quality and recognition of prior work needs to trump openness. That seems to me to be a foundational principle of academic integrity.]
I am proposing that the publisher ping the authors of cited sources which are not openly accessible with some version of the following message:
- Good news! We are about to publish a peer-reviewed article that cites your article such-and-such. This is a tribute to your good work.
- We notice that your article is not openly accessible, yet could be.
- Here’s how to do it: xxx-yyy-zzz.
- Did you know that if your article is openly accessible not only will it be read more times but it is also more likely to be cited in the future? [I understand that empirical research backs up this assertion.]
- If you have any questions let us know.
- Keep up the good work.
In cases where there are multiple authors of a cited source I think the publisher should ping all of the authors even though in some cases it’s only the lead author who is recognized to have the self-archive rights. This is because the lead author may already have tenure and therefore may not be as strongly self-interested in further recognition. But her co-authors are very likely to be her graduate students and post-docs and protégés. Perhaps some social pressure will help nudge the lead author to exercise their self-archival rights.
As one senior executive of a large scholarly press pointed out to me in discussing this idea, communicating to authors of cited sources is an opportunity to strengthen their brand with important authors in the field. So the self-interest for the publisher in doing this encouragement of opening the cited sources is more than just to improve the quality of the new article; it’s also about establishing a presence with potential future contributors to their journals.
Publishers who are purely open access publishers have another motivation to undertake this pinging process. They are not just scholarly publishers — they also have a mission to actively advocate for open access. Pinging these authors of cited sources is a great opportunity to educate broadly among scholars in the disciplines that they publish in. And answering questions that come back from these authors can be an important measure of how well scholars understand the plethora of issues that continually arise in the industry.
Another interesting effect of this pinging proposal is that pinging will tend to gang up on the most important articles since they will be the ones that are more often cited.
What’s Wrong with this Proposal?
I always remember the advice of an old friend, Gerry Weinberg, from his book Are Your Lights On?, “If you can’t think of at least three things wrong with your proposal you probably don’t understand the problem.” So here are some things potentially wrong, many of them raised by people I’ve shopped this idea with recently:
- While it is clear that articles with cited sources that can be clicked through to access the full text are better, is the cost worth the benefit?
- If not, are there enough clear ways to automate much of the process so that the costs go down?
- Is a publisher likely to have access to the contact information for a sufficient number of the cited sources to make this a useful effort?
- Are changes needed to citation standards, link-resolvers, DOIs or other things that effect a user’s ability to locate cited sources which may reside in different places?
- What about cited sources where the author is no longer living or accessible?
Let’s First Shine a Light on This and See What that Evokes
1) I think we need a simple tool that will shine a light on the accessibility of cited sources. My non-technical description would be a tool which could:
- Ingest a reference list, a bibliography, or even just an author’s c.v.
- It would provide a score (%) of referenced sources which are not accessible and would highlight them in the output.
2) This could help automate the process of identifying which authors to ping for publishers taking up that process.
3) It could also be used by individual scholars who are good O.A. citizens and have decided to go through their whole set of published work and share any that have not been archived yet.
4) Run on a grander scale such a tool could be used by the editor of a journal to score the journal on % of cited sources which are openly accessible.
5) Similarly, an institution which wanted to support open access could score the total output of the university against this measure.
6) Or perhaps even a discipline which took on the mission of having its scholarly literature open available to all.
7) Various advocacy organizations could use a tool like this to assess which articles are referenced the most in some large body of work and focus their advocacy on the authors of the most popular yet still inaccessible articles.
8) Everything said here about publishing a journal article could equally be said about a couple of other content types important to academic publishing, namely: annotated bibliographies and curated lists of resources. Again, inclusion of a cited source that is open is a better user experience of that bibliography and is also a moment of good news to the author that their work is now being recognized. A bibliography where most or all of the cited sources are available is clearly of greater use than one which simply identifies sources which are inaccessible.
I think that the combination of these measures could definitely “move the needle” in addressing misunderstandings among scholars about open access and begin to overcome the inertia that otherwise holds back the full realization of the benefits of open access to the world’s scientific and scholarly enterprise. Keeping an eye on the total utility of a journal article to the least resourced scholar (other than access to the Internet) I think we can deploy continuous improvement until the full benefits of open access for scholarly journals is achieved.
Feedback so Far
I’ve now shopped around this basic idea to a wide variety of people including senior management of a couple significant academic publishers, scholars (among friends and family) in a wide variety of disciplines including STEM, social sciences, and humanities, well recognized experts in Open Access, Alt Metrics, members of NISO, librarians, and open access publishers. Almost everyone so far has found the idea intriguing and worth further examination. One put it this way, “I’m not aware of publishers doing this today and I think it’s definitely something they should try.” Another said, “This is really good. It’s simple and I’m not aware of anyone paying attention to this.”
I’ve now found two publishers, one an OA publisher and another a subscription publisher who are pinging authors of works referred to in the publisher’s new publications. Neither one is doing this to ping authors about sharing their articles. Rather, they are pinging authors of cited works in order to strengthen the connection between the publisher’s new authors and the authors of referenced works. These two publishers both say that it’s a lot of work, but they nevertheless stand as a bit of an “existence proof” that it can be done and that it can be motivated by the self-interest of publishers.
I would like to single out Peter Suber’s Book, Open Access, published in 2012 by MIT Press (https://mitpress.mit.edu/books/open-access and now available open access) which was especially helpful to me to clarify my ideas about these suggestions.